伺服器版本 | docker軟體版本 | Kubernetes(k8s)叢集版本 | CPU架構 |
---|---|---|---|
CentOS Linux release 7.4.1708 (Core) | Docker version 20.10.12 | v1.21.9 | x86_64 |
Kubernetes叢集架構:k8scloude1作為master節點,k8scloude2,k8scloude3作為worker節點
伺服器 | 作業系統版本 | CPU架構 | 程序 | 功能描述 |
---|---|---|---|---|
k8scloude1/192.168.110.130 | CentOS Linux release 7.4.1708 (Core) | x86_64 | docker,kube-apiserver,etcd,kube-scheduler,kube-controller-manager,kubelet,kube-proxy,coredns,calico | k8s master節點 |
k8scloude2/192.168.110.129 | CentOS Linux release 7.4.1708 (Core) | x86_64 | docker,kubelet,kube-proxy,calico | k8s worker節點 |
k8scloude3/192.168.110.128 | CentOS Linux release 7.4.1708 (Core) | x86_64 | docker,kubelet,kube-proxy,calico | k8s worker節點 |
Kubernetes(k8s)資料卷volumes型別眾多,本文介紹資料卷volumes之一NFS資料卷。
使用資料卷volumes的前提是已經有一套可以正常執行的Kubernetes叢集,關於Kubernetes(k8s)叢集的安裝部署,可以檢視部落格《Centos7 安裝部署Kubernetes(k8s)叢集》https://www.cnblogs.com/renshengdezheli/p/16686769.html
nfs 卷能將 NFS (網路檔案系統) 掛載到你的 Pod 中。 不像 emptyDir 那樣會在刪除 Pod 的同時也會被刪除,nfs 卷的內容在刪除 Pod 時會被儲存,卷只是被解除安裝。 這意味著 nfs 卷可以被預先填充資料,並且這些資料可以在 Pod 之間共用。
說明:在使用 NFS 卷之前,你必須運行自己的 NFS 服務器並將目標 share 匯出備用。
還需要注意,不能在 Pod spec 中指定 NFS 掛載可選項。 可以選擇設定伺服器端的掛載可選項,或者使用 /etc/nfsmount.conf。 此外,還可以通過允許設定掛載可選項的持久卷掛載 NFS 卷。
此次共用儲存以nfs為例,在一臺機器上安裝NFS伺服器端,k8s的兩個worker安裝NFS使用者端。
etcd1機器作為NFS的伺服器端,安裝NFS
[root@etcd1 ~]# yum -y install nfs-utils
[root@etcd1 ~]# rpm -qa | grep nfs
libnfsidmap-0.25-19.el7.x86_64
nfs-utils-1.3.0-0.68.el7.2.x86_64
啟動NFS
#使nfs開機自啟動並現在就啟動
[root@etcd1 ~]# systemctl enable nfs-server --now
Created symlink from /etc/systemd/system/multi-user.target.wants/nfs-server.service to /usr/lib/systemd/system/nfs-server.service.
#檢視nfs狀態
[root@etcd1 ~]# systemctl status nfs-server
● nfs-server.service - NFS server and services
Loaded: loaded (/usr/lib/systemd/system/nfs-server.service; enabled; vendor preset: disabled)
Active: active (exited) since 二 2022-01-18 17:24:24 CST; 8s ago
Process: 1469 ExecStartPost=/bin/sh -c if systemctl -q is-active gssproxy; then systemctl reload gssproxy ; fi (code=exited, status=0/SUCCESS)
Process: 1453 ExecStart=/usr/sbin/rpc.nfsd $RPCNFSDARGS (code=exited, status=0/SUCCESS)
Process: 1451 ExecStartPre=/usr/sbin/exportfs -r (code=exited, status=0/SUCCESS)
Main PID: 1453 (code=exited, status=0/SUCCESS)
CGroup: /system.slice/nfs-server.service
1月 18 17:24:24 etcd1 systemd[1]: Starting NFS server and services...
1月 18 17:24:24 etcd1 systemd[1]: Started NFS server and services.
建立NFS共用目錄,並把目錄/sharedir共用出去
#建立/sharedir作為共用目錄
[root@etcd1 ~]# mkdir /sharedir
[root@etcd1 ~]# vim /etc/exports
#把/sharedir目錄共用出去
[root@etcd1 ~]# cat /etc/exports
/sharedir *(rw,async,no_root_squash)
[root@etcd1 ~]# exportfs -arv
exporting *:/sharedir
在k8s叢集的worker節點安裝nfs的使用者端
[root@k8scloude3 ~]# yum -y install nfs-utils
#安裝nfs的使用者端
[root@k8scloude2 ~]# yum -y install nfs-utils
檢視etcd1(192.168.110.133)機器共用出來的目錄是哪個?
[root@k8scloude2 ~]# showmount -e 192.168.110.133
Export list for 192.168.110.133:
/sharedir *
把192.168.110.133:/sharedir的目錄掛載到/mnt
[root@k8scloude2 ~]# mount 192.168.110.133:/sharedir /mnt
[root@k8scloude2 ~]# df -hT /mnt
檔案系統 型別 容量 已用 可用 已用% 掛載點
192.168.110.133:/sharedir nfs4 150G 2.5G 148G 2% /mnt
設定nfs卷,指定共用資料卷的型別為nfs,指定NFS伺服器IP和共用目錄
[root@k8scloude1 volume]# vim share-nfs.yaml
[root@k8scloude1 volume]# cat share-nfs.yaml
apiVersion: v1
kind: Pod
metadata:
creationTimestamp: null
labels:
run: hostpath
name: nfsshare
spec:
#nodeName指定該pod執行在k8scloude2節點
nodeName: k8scloude2
terminationGracePeriodSeconds: 0
volumes:
- name: v1
#資料卷的型別為nfs
nfs:
#nfs伺服器地址
server: 192.168.110.133
#共用目錄
path: /sharedir
#readOnly: true唯讀
#readOnly: true
containers:
- image: nginx
imagePullPolicy: IfNotPresent
name: h1
resources: {}
volumeMounts:
- name: v1
#把v1卷掛載到/xx目錄
mountPath: /xx
dnsPolicy: ClusterFirst
restartPolicy: Always
status: {}
建立pod
[root@k8scloude1 volume]# kubectl apply -f share-nfs.yaml
pod/nfsshare created
[root@k8scloude1 volume]# kubectl get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nfsshare 1/1 Running 0 3s 10.244.112.189 k8scloude2 <none> <none>
進入pod
[root@k8scloude1 volume]# kubectl exec -it nfsshare -- bash
root@nfsshare:/# ls /xx/
#往共用目錄裡寫入資料
root@nfsshare:/# echo "well well well" >/xx/log.txt
root@nfsshare:/#
root@nfsshare:/# exit
exit
k8scloude2機器上有對應的檔案
[root@k8scloude2 ~]# cat /mnt/log.txt
well well well
etcd1機器上也有對應的檔案
[root@etcd1 ~]# cat /sharedir/log.txt
well well well
刪除pod
[root@k8scloude1 volume]# kubectl delete pod nfsshare --force
warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "nfsshare" force deleted
[root@k8scloude1 volume]# kubectl get pods -o wide
No resources found in volume namespace.
設定nfs卷,指定共用資料卷的型別為nfs,指定NFS伺服器IP和共用目錄,不過讓pod執行在k8scloude3上
[root@k8scloude1 volume]# vim share-nfs.yaml
#nodeName: k8scloude3 指定pod執行在k8scloude3上
[root@k8scloude1 volume]# cat share-nfs.yaml
apiVersion: v1
kind: Pod
metadata:
creationTimestamp: null
labels:
run: hostpath
name: nfsshare
spec:
nodeName: k8scloude3
terminationGracePeriodSeconds: 0
volumes:
- name: v1
nfs:
server: 192.168.110.133
path: /sharedir
containers:
- image: nginx
imagePullPolicy: IfNotPresent
name: h1
resources: {}
volumeMounts:
- name: v1
mountPath: /xx
dnsPolicy: ClusterFirst
restartPolicy: Always
status: {}
建立pod
[root@k8scloude1 volume]# kubectl apply -f share-nfs.yaml
pod/nfsshare created
[root@k8scloude1 volume]# kubectl get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nfsshare 1/1 Running 0 3s 10.244.251.251 k8scloude3 <none> <none>
因為使用的是NFS共用儲存卷,進入pod,對應的目錄都有檔案
[root@k8scloude1 volume]# kubectl exec -it nfsshare -- bash
root@nfsshare:/# cat /xx/log.txt
well well well
root@nfsshare:/#
root@nfsshare:/# exit
exit
當pod排程在k8scloude3上,也出現了相應NFS掛載
[root@k8scloude3 ~]# df -h | grep 192.168.110.133
192.168.110.133:/sharedir 150G 2.5G 148G 2% /var/lib/kubelet/pods/4ebc5f6d-e13c-4bea-a323-3067c4a6e966/volumes/kubernetes.io~nfs/v1
刪除pod
[root@k8scloude1 volume]# kubectl delete pod nfsshare --force
warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "nfsshare" force deleted
[root@k8scloude1 volume]# kubectl get pods -o wide
No resources found in volume namespace.
當刪除pod,則這個掛載消失
[root@k8scloude3 ~]# df -h | grep 192.168.110.133
NFS卷存在的問題:每個人都可能連線到儲存伺服器,都必須使用root許可權,對伺服器來說具有安全隱患,還要花時間學習NFS知識。